Advanced Techniques in Pediatric Abdominopelvic Oncologic Magnetic Resonance Imaging

A d v a n c e d Te c h n i q u e s i n Pe diatric Abdomi no p elvic Oncologic Magnetic Resonance Imaging Ethan A. Smith, MD KEYWORDS  Oncology  Malignancy  PET-MR imaging  Whole-body imaging  Dynamic contrast enhancement  Children

KEY POINTS  The use of magnetic resonance imaging (MRI) in the setting of pediatric abdominopelvic malignancy is increasing because of concerns regarding ionizing radiation exposure, superior contrast resolution compared with computed tomography (CT), and recent technical advances.  Many ongoing oncologic clinical trials now incorporate MR imaging in their protocols as an alternative to CT imaging.  Advanced MR imaging techniques, such as diffusion-weighted imaging, dynamic contrast– enhanced imaging (MR imaging perfusion), and 18F-fluorodeoxyglucose positron emission tomography MR imaging, provide functional assessment of abdominopelvic tumors in addition to morphologic information.  Whole-body (WB) MR imaging has been shown to be useful in staging and surveillance of pediatric malignancy without the use of ionizing radiation, as is required for other currently used WB imaging techniques. The major disadvantage of WB-MR imaging is its lack of specificity compared with more targeted imaging strategies, such as metaiodobenzylguanidine (MIBG) scanning in neuroblastoma.  Major weaknesses of MR imaging in pediatric oncologic imaging include difficulty evaluating the lung parenchyma for metastatic disease and detection of calcification within tumors.

Magnetic resonance imaging (MRI) is firmly established as a useful imaging technique in pediatric patients. However, MRI use has lagged behind the use of computed tomography (CT) in pediatric abdominopelvic oncology because of a variety of factors, including issues related to cancer research protocols, concerns about the sensitivity of MR imaging, and a general lack of awareness among clinical providers as to the feasibility and advantages of current MR imaging techniques.

As abdominopelvic MRI has become more widely available, pulse sequences have become faster and more robust, and image quality has improved, the use of MRI in pediatric oncologic imaging has increased. This growth in MRI use in most instances has been as an alternative to CT imaging for morphologic evaluation, although recent technologic advances now allow functional assessment of abdominopelvic tumors as well. This article focuses on advanced MRI techniques currently being clinically used or investigated in pediatric abdominopelvic oncologic imaging. A range

Section of Pediatric Radiology, Department of Radiology, C.S. Mott Children’s Hospital, University of Michigan Health System, 1540 E. Hospital Dr., SPC 4252, Ann Arbor, MI 48109-4252, USA E-mail address: [email protected] Magn Reson Imaging Clin N Am 21 (2013) 829–841 http://dx.doi.org/10.1016/j.mric.2013.06.002 1064-9689/13/$ – see front matter Ó 2013 Elsevier Inc. All rights reserved.

mri.theclinics.com

INTRODUCTION

830

Smith of different techniques, including diffusion-weighted imaging (DWI), dynamic contrast-enhanced (DCE) MRI, whole-body (WB) MR imaging, and 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography MR (PET-MRI) imaging, are discussed. In addition, the increasing acceptance of MRI in pediatric cancer research protocols and a brief discussion of specific issues related to performing MR imaging in children, such as sedation, are addressed.

PROTOCOL AND PATIENT ISSUES A standardized, but flexible, MRI protocol for abdominopelvic tumor imaging is important for success in pediatric oncology imaging. Standardization of the protocol, including standardized pulse sequence parameters and imaging planes, allows for signal intensity and size measurements that can be used for characterization of the lesion(s) at the time of initial diagnosis and for assessing treatment response on follow-up imaging. A large number of pediatric oncology patients, especially those with solid tumors, are formally treated and followed after therapy using specific protocols, often through the Children’s Oncology Group (COG). Having a standardized, reproducible protocol is important in these patients because imaging findings commonly determine patient eligibility for clinical trials and changes in lesions over time directly affect patient treatment and follow-up decisions. Fortunately, most COG protocols are straightforward and commonly used clinical pediatric abdominopelvic MR imaging protocols often fulfill necessary imaging requirements. Our institutional routine clinical abdominopelvic mass MRI protocol is presented in Table 1. MRI protocols can be altered in order to optimize evaluation depending on the clinical setting and suspected tumor type. For example, in the setting of a known or suspected primary liver neoplasm, we routinely perform dynamic postcontrast MR imaging through the liver, including arterial phase imaging. In this clinical situation we also routinely administer a hepatocyte-specific contrast agent (gadoxetate disodium; Eovist, Bayer HealthCare, Wayne, NJ) and acquire additional delayed hepatocyte phase imaging at 10 and 20 minutes following injection. Tailoring the MRI protocol to the tumor or organ of interest often provides added information, while maintaining the ability to perform routine assessments of signal intensity and size. The use of sedation and general anesthesia is commonplace in young children undergoing

abdominopelvic MRI for the assessment of suspected or known malignancy, because of several factors. Most young children are unable to remain sufficiently motionless for MRI, because imaging times are commonly in the 30 to 60 minute range. In addition, young children often have difficultly cooperating with breath-holding instructions. The physical appearance of the MRI scanner and the noises produced during imaging may also be frightening to some children, making cooperation even less likely.1 To this end, the use of ear plugs, noise-cancelling headphones, and distraction devices (eg, music headphones, movie goggles) may be helpful, especially in school-aged children.2 Preprocedure preparation with play therapy and other strategies have also been shown to be effective in reducing the need for general anesthesia.1,3 Imaging studies performed under general anesthesia have been shown to require longer recovery times and are more costly than those performed without general anesthesia.1 Some studies have shown that repeated exposure to general anesthetics may be associated with negative long-term cognitive effects, although others studies have not found such an association.4,5 When deciding between the use of nonsedated CT and MRI performed under sedation or general anesthesia, the advantages and disadvantages of both modalities should be carefully considered, including the potential harmful effects of both ionizing radiation and general anesthesia.

DIFFUSION-WEIGHTED IMAGING DWI uses the motion of water molecules to provide image contrast and to characterize tissues. The motion of water molecules in a glass of water is random (Brownian motion), whereas in biological tissues this random motion can be impeded (or restricted), primarily by cell membranes.6 The first widespread clinical use of DWI was in neuroimaging, specifically stroke imaging, because the cellular swelling that accompanies acute infarction (cytotoxic edema) results in restricted water motion and signal hyperintensity. DWI is now being evaluated as an oncologic imaging biomarker for distinguishing benign from malignant lesions and for evaluating response to therapy, including in the setting of pediatric abdominopelvic malignancy. The increased cellularity of many malignant tumors compared with normal tissues impedes the motion of water and presents as restricted diffusion (signal hyperintensity) on DWI.7 In current clinical practice, DWI sequences (with higher b values) are used for qualitative assessment, because tumors and abnormal lymph nodes are usually

Pediatric Abdominopelvic Oncologic MR Imaging

Table 1 Abdominopelvic mass MR imaging protocol at 1.5 T

Plane Slice thickness TR (ms) TE (ms) Respiratory compensation NSA Fat saturation b values

T1W TSE

T2W TSE

T2W SSFSE

T2W TSE

DWI EPI

Coronal 5 554 4.6 None

Coronal 5 Shortest 80 Trigger

Coronal 5 Shortest 80 Trigger

Axial 5 Shortest 80 Trigger

Axial 6 Shortest Shortest Trigger

4 N —

2 Y —

1 N —

2 Y —

3 Y 0, 100, 750

T1W In-phase T1W 3D FFE T1W 3D FFE FFE Precontrast Postcontrast Axial 5 Shortest 4.6 Breath hold 1 N —

Axial 2-3 Shortest Shortest Breath hold

3 planes 2-3 Shortest Shortest Breath hold

1 Y —

1 Y —

Shortest means the shortest time required as determined by scanner. Abbreviations: 3D, three-dimensional; EPI, echo planar imaging; FFE, fast field echo; Free, free breathing; IR, inversion recovery; NSA, number of signal averages; SSFSE, single-shot fast spin-echo; STIR, Short tau inversion recovery; T1W, T1 weighted; T2W, T2 weighted; TE, echo time; TR, repetition time; Trigger, respiratory triggered; TSE, turbo spin-echo.

hyperintense relative to the background, allowing increased sensitivity (Fig. 1). The degree of restricted diffusion can be quantified by performing DWI with multiple b values and calculating a

map of apparent diffusion coefficient (ADC) values (Fig. 2). The number of b values required (eg, 2, 3, 4 or more) as well as their absolute values (0 to greater than 1000 m/s2) have yet to be established

Fig. 1. A 15 -year-old boy with nasopharyngeal carcinoma, status post therapy. (A) Axial fused FDG-PET-CT images show focal, abnormal hypermetabolism in the spleen and liver (arrows). (B) Axial T2W fat-saturated MR image showing focal lesions in the liver and spleen (arrows) corresponding with the PET avid lesions. (C) Axial DWI (b 5 800) shows hyperintense signal corresponding with the liver lesion (arrow). Several additional lesions are also present in the liver, none of which were visible on the other sequences or PET-CT (arrowheads). The splenic lesion is relatively hypointense relative to the normally hyperintense spleen (arrow). (D) Axial T1W spoiled gradient recalled echo (SPGR) 20 minutes after intravenous administration of gadoxetate disodium (Eovist) shows hypointensity of the larger liver lesion, consistent with a metastasis (arrow).

831

832

Smith

Fig. 2. A 5-year-old girl with abdominal distention caused by Burkitt lymphoma. The patient underwent MR imaging because of suspected peritoneal and omental masses based on initial ultrasound imaging. (A) Axial T2W fat-suppressed image showing a large omental mass (arrows) and a small amount of ascites. (B) Axial diffusionweighted image (b 5 750) shows a large omental mass that appears hyperintense because of restricted diffusion (arrows). (C) The mass appears hypointense on the corresponding ADC image (arrows), confirming restricted diffusion. (D) Axial fused 18F-FDG-PET-CT image performed the following day confirms that the large omental mass is hypermetabolic (arrows).

for accurate, reproducible ADC calculation in pediatric oncologic imaging. Humphries and colleagues8 evaluated ADC values in 19 children with extracranial soft tissue masses. Although they found an inverse relationship between the ADC value and the cellularity of tissues on histopathology, there was no significant difference between benign and malignant lesions, and their final conclusion was that ADC value alone could not accurately distinguish between benign and malignant tumors. Contradictory results were observed in a study performed by Kocaoglu and colleagues,7 who found that significantly lower ADC values were observed with malignant lesions in a group of 31 pediatric abdominal masses (15 benign, 16 malignant), although there was some overlap between the two groups. Interestingly, these studies used different methods for determining the region of interest for calculation of the mean ADC values.7,8 More recently, Gahr and colleagues9 retrospectively evaluated 19 pediatric neurogenic tumors and found a significant difference between the ADC values of neuroblastoma and ganglioneuroblastoma/ganglioneuroma. Their

conclusion was that neuroblastoma could be differentiated from the other less aggressive neurogenic tumors based on ADC values, and that the lower ADC values observed in neuroblastoma were caused by the greater cellularity of this neoplasm. In a feasibility study involving 7 pediatric oncology patients with 9 tumors, McDonald and colleagues10 analyzed ADC values before and after completion of chemotherapy. The ADC values of all tumors changed following chemotherapy, with an increase in the median ADC values seen in most lesions. The increased ADC value (indicating less impeded diffusion and less hyperintense signal on DWI images) was thought to represent decreased cellularity of the tumor, thus representing treatment response. The largest increases in ADC value were observed in those tumors that showed the greatest response to therapy on histopathologic evaluation (Fig. 3). WB-DWI has recently become available on state-of-the-art MR imaging scanners. This technique can be used to complement standard oncologic MR imaging protocols or WB-MR imaging examinations, increasing the conspicuity of small

Pediatric Abdominopelvic Oncologic MR Imaging

Fig. 3. A 15 year-old boy with recurrent rhabdomyosarcoma. (A) Axial T1-weighted postcontrast image shows a heterogeneously enhancing left-sided abdominal mass (arrows). (B) Axial diffusion-weighted image (b 5 800, 1.5-T) at the same level shows increased signal within the mass (arrows), consistent with restricted diffusion. (C) Axial diffusion-weighted image (3-T) after chemotherapy shows slightly decreased size of the mass. The mass also appears less hyperintense and more heterogeneous (arrows). (D) Axial fused 18F-FDG-PET-CT image performed the same day shows only a peripheral rim of 18F-FDG avidity (arrows). Most of the mass is hypometabolic because of presumed treatment effect. Subsequent surgical resection of the mass and histopathologic evaluation showed that approximately 90% of the mass was necrotic.

lesions and possibly examination sensitivity.11 WB-DWI can now be performed in a reasonable amount of time (20–30 minutes).12 WB-DWI is performed using a floating table technique, similar to other WB-MR imaging techniques (such as MR angiography), and a high b value is generally chosen (eg, 800–1000 m/s2) in order to maximally suppress nonpathologic background signal from the body. One WB-DWI technique that can be performed with the patient free breathing is called DWI with background body suppression (DWIBS).13 Acquired gray-scale images can then be inverted to resemble 18F-FDG-PET images, and three-dimensional (3D) maximum intensity projections can be created to allow assessment of the entire body on a single image (Figs. 4 and 5).13 Kwee and colleagues14 compared WB-MR imaging, WB-DWI, and standard CT in the staging of lymphoma and found that a combination of WB-MR imaging and WB-DWI increased the stage at initial diagnosis of 4 of 28 patients compared with CT.

WHOLE-BODY MAGNETIC RESONANCE IMAGING WB-MR imaging can be used to evaluate for sites of neoplasm remote from the primary tumor for staging purposes as well as to screen patients determined to be at increased risk for malignancy because of certain genetic abnormalities (eg, Li-Fraumeni syndrome). Commonly used techniques for WB (or near-WB) imaging in the setting of malignancy include radiographic skeletal surveys, CT, and a variety of nuclear medicine studies, such as technetium-99 m methylene diphosphonate (Tc-99 m MDP) bone, octreotide, metaiodobenzylguanidine (MIBG), and 18F-FDGPET scans, all of which have the disadvantage of using ionizing radiation. Several studies have shown that WB-MR imaging is useful for detecting both skeletal and extraskeletal metastases in pediatric patients.15–18 WBMR imaging can be performed as a stand-alone examination, or WB sequences can be added to other standard MR

833

834

Smith

Fig. 4. A 12-year-old boy with B-cell non-Hodgkin lymphoma. (A) Coronal WB STIR image shows enlarged, hyperintense cervical and mediastinal lymph nodes (arrows). The spleen is enlarged. (B) 3D maximum intensity projection WB DWIBS shows widespread cervical, thoracic, and abdominopelvic pathologic lymphadenopathy (arrows). The spleen is enlarged and restricts diffusion. (C) 3D maximum intensity projection 18F-FDG-PET image obtained after MR imaging and before treatment also shows widespread pathologic lymphadenopathy (arrows). The spleen is enlarged and hypermetabolic. The DWIBS shows a decreased number of abnormal mediastinal lymph nodes compared with the 18F-FDG-PET image, in part secondary to obscuration by the overlying the spine. (Courtesy of Drs Thomas Kwee and Annemieke Littooij, Utrecht, Netherlands.)

protocols used to stage and follow-up pediatric oncology patients.17,19,20 Several advances in technology have made WB-MR imaging feasible when using state-ofthe-art scanners, including floating table

capabilities, improved body coils (both integrated into the scanner and surface coils), and faster imaging techniques.20 The most commonly used pulse sequence in WB-MR imaging is short tau inversion recovery (STIR), chosen because of its Fig. 5. A 12-year-old boy with B-cell non-Hodgkin lymphoma (same patient as in Fig. 4). Posttreatment WB short tau inversion recovery (STIR) image (A) and 3D maximum intensity reconstructed WB DWIBS (B) both show significant treatment response, with decreased size and number of abnormal lymph nodes. The spleen remains mildly enlarged. (Courtesy of Drs Thomas Kwee and Annemieke Littooij, Utrecht, Netherlands.)

Pediatric Abdominopelvic Oncologic MR Imaging sensitivity to signal from fluid (including edema, inflammation, and masses with substantial water content), relative resiliency to artifacts, and homogeneous fat saturation.20–22 This pulse sequence takes advantage of the fact that most pathological tissues in the human body cause T2 prolongation causing conspicuous signal hyperintensity.18,20,22 Additional pulse sequences that have been used for WB-MR imaging include T1-weighted fast spin-echo (FSE) without fat saturation (primarily used for evaluation of bone marrow), twodimensional balanced steady-state free precession, T2-weighted single-shot FSE (SSFSE), and postcontrast 3D gradient recalled echo imaging.20 Most investigators advocate imaging in the coronal plane because of the smaller number of images required to cover the body (compared with the axial and sagittal planes), although some

also perform additional sagittal imaging to increase sensitivity for lesion detection, especially in the spine, ribs, and sternum (Fig. 6).18,20,22 WB-MR imaging generally is not performed in the axial plane because of the large number of images and amount of time required to cover the body.22 Multiple prior investigations have examined the usefulness of WB-MR imaging in the work-up of both adult and pediatric oncology patients. Solid tumors, such as neuroblastoma and other tumors that have a propensity for osseous metastases, have been most widely studied. The staging of lymphoma has been studied as well. In addition, WB-MR imaging has been evaluated as a screening tool for patients with certain genetic familial cancer syndromes, in some nonneoplastic conditions (eg, Langerhans cell histiocytosis), and

Fig. 6. WB-MR imaging examination in a 9-year-old girl with Li-Fraumeni syndrome and history of rhabdomyosarcoma. Coronal STIR (A) and T1-weighted turbo spin-echo (B) images show portions of brain, lungs, abdominopelvic viscera, spine, and extremities. No abnormalities are seen.

835

836

Smith in certain clinical presentations of suspected infection.20 In a pilot study in 2002, Mazumder and colleagues16 found that coronal STIR WB-MR imaging performed similarly to standard WB imaging techniques in the metastatic work-up in a small group of pediatric patients with small round cell malignancies. Laffan and colleagues17 added coronal and sagittal STIR WB imaging to a group of 10 pediatric patients with suspected malignancy, but in whom a pathologic diagnosis was not known. They found distant sites of neoplasm in 8 of 10 patients and concluded that WB-MR imaging may be useful in the initial work-up of suspected malignancy.17 Goo19 found that WBMR imaging was more sensitive than both Tc-99 m MDP bone scan and MIBG in a group of patients with neuroblastoma, although the positive predictive value of WB-MR imaging for metastasis was less than that of MIBG. However, this study was limited by the lack of a standardized WB-MR imaging protocol. More recently, in a large multi-institutional study sponsored by the American College of Radiology Imaging Network (ACRIN) evaluating WB-MR imaging in pediatric patients with known malignancy, Seigel and colleagues23 were unable to show noninferiority criteria for accuracy of WB-MR imaging compared with conventional imaging methods when evaluating a combination of lymphoma, neuroblastoma, and soft tissue sarcoma. This study did find that, if lymphoma was excluded from the analysis, the diagnostic accuracy of WB-MR imaging was improved, and that it performed better than conventional imaging in the evaluation of osseous metastatic disease.23 WB-MR imaging does have known limitations. The most important limitation of this technique is its reduced specificity compared with more targeted imaging strategies, such as MIBG scanning in neuroblastoma. A variety of nonneoplastic processes can cause hyperintense signal on STIR images and may be confused with metastatic neoplasm, including benign infectious and inflammatory processes.20,21 Normal red marrow shows mildly hyperintense signal on STIR images and mildly hypointense signal on T1-weighted images, and potentially can be confused for a pathologic lesion if the reader is not familiar with this appearance. Large amounts of red marrow can be present in young pediatric patients, in patients receiving certain hematopoietic medications (eg, granulocyte-macrophage colony stimulating factor), and in patients undergoing red marrow reconversion as a response to prior chemotherapy.21 Compared with CT, both standard MR imaging and WB-MR imaging have decreased sensitivity

for detecting lung parenchymal abnormalities, including lung nodules.24 Both normal and pathologic lymph nodes are hyperintense on STIR imaging, which means that, unlike 18F-FDG-PET in which hypermetabolism is used as the primary criterion for detecting abnormal lymph nodes, the diagnosis of abnormal lymph nodes on WB-MR imaging still depends on morphology and imaging size criteria.21 Tumoral calcifications, such as those frequently encountered in patients with neuroblastoma, often appear hypointense on both T1-weighted and STIR images, sometimes making them difficult to visualize on WB-MR imaging.19 In addition, given the large fields of view and section thicknesses (several millimeters) used to accomplish WB-MR imaging, the sensitivity for detecting very small lesions (ie, less than 1 cm in diameter) may be diminished compared with standard MR imaging (particularly in the absence of simultaneous WB-DWI).20,21

DYNAMIC CONTRAST-ENHANCED MAGNETIC RESONANCE IMAGING DCE-MRI describes a group of techniques that attempt to assess angiogenesis and perfusion in order to characterize the microvascular environment of tumors. Malignant neoplasms have been shown to have disorganized, functionally abnormal vasculature compared with normal tissues, thought to be in part secondary to an imbalance of angiogenic factors.25 Initial studies in breast carcinoma revealed that permeability of contrast agents within malignant neoplasms is altered compared with normal tissues because of relative microvascular disorganization within the tumor.26,27 Three different methods using different types of contrast agents have been investigated to exploit the differences in microvasculature between tumors and normal tissues: (1) macromolecular contrast agents (eg, labeled albumin) that take advantage of the increased permeability of tumor vessel basement membranes to allow larger molecules (which would normally be confined to the vascular space) to diffuse into the tumor; (2) contrast agents designed to pool at sites of angiogenesis; and (3) low-molecular-weight contrast agents.28,29 The remainder of this section focuses on low-molecular-weight gadolinium-based contrast agents, because these are widely available and currently in use in clinical practice. The kinetic properties of low-molecular-weight gadolinium-based contrast agents primarily depend on 3 factors: tissue perfusion, vessel wall permeability, and the rate of diffusion.28 Tissue perfusion determines how much of the contrast agent is delivered to the tumor and depends on

Pediatric Abdominopelvic Oncologic MR Imaging blood flow. Once the contrast agent arrives at the tissue, the bolus is initially confined within the vascular space, with a variable amount rapidly passing into the extravascular space. Because low-molecular-weight contrast agents do not cross cell membranes and enter cells, the site of distribution is most accurately defined as the extracellular extravascular space (EES), or ve.28,30 The rate of contrast passage into this space is influenced by multiple factors, including tissue perfusion, the permeability of the blood vessels within the tissue, and the surface area available.28,31–33 These factors are described by the transfer constant, or Ktrans, for a tissue.30 A third variable, the rate constant, or kep, describes the flux of contrast agent between the EES (ve) and the vascular space, and is proportional to the transfer constant (kep 5 Ktrans/ve) (Table 2).30 As the intravascular volume of contrast agent is depleted (both by distribution and excretion), contrast agent begins to diffuse from the EES back into the vasculature and eventually is excreted from the body (primarily through the kidneys, and possibly the liver, depending on the contrast agent). This washout of contrast agent may be accelerated if the permeability of a tissue’s vasculature is increased.28 DCE-MRI has been most extensively studied in breast cancer, but has been evaluated in numerous other oncologic (eg, genitourinary and musculoskeletal tumors) and inflammatory conditions (eg, Crohn disease) as well. Although both T1-weighted and T2-weighted imaging can be used for DCE-MR imaging, most studies have focused on T1-weighted imaging, exploiting the T1 shortening that occurs after the intravascular

Table 2 Definitions of kinetic factors commonly used in dynamic contrast-enhanced MR imaging Name K

trans

kep

ve

Units Definition

Transfer min constant

1

Volume transfer constant between blood plasma and EES Rate min 1 Rate constant constant between EES and plasma volume EES none Volume of extracellular, extravascular space per unit of tissue

Adapted from Tofts PS, Brix G, Buckley DL, et al. Estimating kinetic parameters from dynamic contrastenhanced T1-weighted MRI of a diffusible tracer: Standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–32.

administration of gadolinium-based contrast agents.28,34 In pediatric patients, most reports have described the use of DCE-MR imaging in bone tumors, such as osteosarcoma.35,36 In a series of 31 patients with primary malignant bone tumors, Reddick and colleagues35 found that kep after completion of preoperative (neoadjuvant) chemotherapy was a statistically significant predictor of disease-free survival. More recently, Gou and colleagues36 evaluated a group of 69 patients with nonmetastatic osteosarcoma and found that changes in several of the DCE-MR imaging parameters after preoperative chemotherapy, including Ktrans, kep, and ve, predicted treatment response at histology, event-free survival, and overall survival.36 DCE-MRI has been evaluated in several adult abdominopelvic malignancies (including cervical, prostate, and bladder cancers), but has not been widely studied in pediatric abdominopelvic cancer. There are several challenges to performing DCE-MR imaging in the pediatric abdomen. First, DCE-MR imaging techniques require multiple repeated imaging acquisitions of the same volume of tissue. Within the abdomen, movement of the diaphragm and intra-abdominal structures caused by normal respiration as well as motion secondary to bowel peristalsis make reimaging the same volume of tissue difficult.28 Second, clinical evaluation of several abdominal tumors benefits from characterization that is achieved through imaging an organ (or even the abdomen and/or pelvis) using multiple specific phases of contrast enhancement (eg, evaluation of pediatric hepatocellular carcinoma or other arterially hyperenhancing tumors). Performance of quantitative DCE-MRI may preclude such an evaluation, unless standard postcontrast imaging is performed after DCE-MRI using a second bolus of contrast agent. However, despite these difficulties, DCE-MRI with formal calculation of perfusion parameters is possible in children with abdominopelvic malignancies and may eventually be incorporated in future clinical trials (Fig. 7). The knowledge gained by DCE-MRI may provide novel imaging biomarkers for establishing benign from malignant tumors, response to medical therapy, and possibly prognosis at the time of initial diagnosis.

POSITRON EMISSION TOMOGRAPHY MAGNETIC RESONANCE IMAGING The use of fused functional and anatomic imaging in the form of PET-CT and single-photon emission SPECT-CT have become cornerstones of oncologic imaging over the past decade in both

837

838

Smith

Fig. 7. Dynamic contrast-enhanced MR imaging in a 3 -year-old boy with a pelvic mass. (A) Free-breathing radial acquisition was performed continuously over 4 minutes during contrast injection. High temporal resolution data (less than 3 seconds) was reconstructed from small number of contiguous spokes (n 5 8) with use of compressed sensing technique called Golden-angle RAdial Sparse Parallel (GRASP). The signal intensity time curve was fitted (B) with the generalized kinetic model to obtain Ktrans (C). The pharmacokinetic modeling was performed using Siemens software Tissue 4D. Population arterial input function was used. (Courtesy of Dr Hersch Chandrana, New York, NY.)

children and adults. In the past, functional imaging, including metabolic imaging with 18F-FDGPET, and anatomic imaging with MR had to be performed in 2 different settings, and then images most often were fused manually or reviewed side by side. Such preliminary forms of PET-MRI were limited primarily to brain imaging, because the rigid skull and fixed landmarks made accurate fusion possible. Fusion of nonbrain PET and MR images, including abdominopelvic images, was more difficult because of a variety of physiologic and anatomic factors.37 Techniques that obtain anatomic MR images in the same setting as PET images have recently become commercially available, creating the possibility of accurate, reliable image fusion (Fig. 8 and 9). At present, there are 2 different paradigms available for PET-MR imaging. Sequential models use separate PET and MR scanners placed in sequence. The two scanners are physically separate, and the patient (on the imaging table) must be moved between the PET scanner and the MR imaging scanner. The main advantage of this technique is that the technology is less complex and is therefore less expensive. The second paradigm

involves integration of the PET and MR imaging into a single imaging unit, eliminating the need to move the patient sequentially between scanners. Although more expensive and technically complex, the main advantage of this technique is improved accuracy of fusion images, because the patient does not have to be physically moved between scanners, decreasing the risk of patient movement and misregistration. There are several potential advantages to using fused PET-MR imaging as opposed to fused PETCT in pediatric oncology patients. First, the inherent soft tissue contrast resolution of MR imaging is superior to both unenhanced and contrastenhanced CT. A second advantage is that, because of its ability to perform DWI, multiphase postcontrast imaging, and imaging with unique intravascular contrast agents, such as gadoxetate disodium (Eovist; Bayer HealthCare), MR imaging can add considerably to the physiologic characterization of tissues. A final key advantage of PET-MR imaging in the pediatric population is that the MRI portion of the examination does not require ionizing radiation, unlike the CT portion of PET-CT. Although there is still exposure to ionizing radiation

Pediatric Abdominopelvic Oncologic MR Imaging

Fig. 8. PET-MR imaging in a 15 -year-old boy with relapsed Hodgkin disease. (A) Coronal T2W SSFSE images show enlarged left axillary and hilar lymph nodes (arrows) as well as pulmonary nodules (arrowhead). (B) Coronal maximum intensity projection PET image and (C) fused PET-MR images both show corresponding abnormal FDG uptake within the lymph nodes (arrows) and pulmonary nodules (arrowhead). (Courtesy of Dr Geetika Khanna, St Louis, MO.)

because of the PET portion of the MR-PET examination, the total dose of imparted radiation is less than with PET-CT. Disadvantages of fused PETMR imaging include the decreased MR imaging

sensitivity for detection of lung nodules and calcifications compared with CT as well as likely increased expense.37,38 Another disadvantage of PET-MR imaging is that many nuclear medicine

Fig. 9. Teenage boy with neurofibromatosis type 1. (A) Coronal T2 weighted image, (B) coronal fused PET-MRI, and (C) PET maximum intensity projection image demonstrate an extensive plexiform neurofibroma involving the abdomen, pelvis and right lower extremity (arrowheads), with minimal FDG uptake. A focal bilobed area of hypermetabolism (arrow) is present. This was considered suspicious, but not diagnostic for malignant transformation, and short term follow up was pursued. (Courtesy of Dr Sarah Milla and Dr Kent Friedman, MD, NYU Langone Medical Center, New York, NY.)

839

840

Smith physicians who currently interpret and report PET-CT studies are not trained or comfortable rendering MR imaging interpretations, which may require PET-MR imaging examinations to be reviewed by both a nuclear medicine physician and a radiologist who routinely interprets clinical MRI examinations. In a review of their early experience with PETMR imaging in both pediatric and adult oncology patients, Antoch and Bockisch38 described the potential advantages and disadvantages of this technique in terms of initial tumor staging. For local disease staging for which the functional/metabolic imaging is less valuable, using MR imaging in place of CT has the advantages of increased soft tissue contrast and improved tissue characterization. In terms of lymph node staging, both CT and MR imaging depend on imaging size criteria and are equivalent with regard to anatomic localization.38 In most instances, both CT and MR imaging are inferior to metabolic imaging using 18 F-FDG-PET for lymph node staging. Evaluation for metastatic disease is a more complex issue. MR imaging is more sensitive than CT and may be more sensitive than 18F-FDG-PET for identification of bone marrow and liver metastases, whereas both 18F-FDG-PET and MR imaging have decreased sensitivity compared with CT for the detection of lung metastases.23,24,38

SUMMARY At present, MRI is an important tool in the evaluation of pediatric abdominopelvic oncologic patients. Going forward, MRI will likely play an increasing role in the management of these patients, including those with abdominopelvic malignancies. The ability of MR imaging protocols to be tailored to optimally evaluate specific suspected or known neoplasms and its lack of ionizing radiation are major advantages compared with existing imaging modalities, particularly CT. Advanced MR imaging techniques, including DWI, DCE-MRI, and PET-MRI, provide functional/physiologic assessment of tumors in addition to morphologic information. Although further research into these applications is needed in the pediatric population, such imaging methods will likely keep MRI at the forefront of pediatric abdominopelvic oncologic imaging for years to come.

REFERENCES 1. Edwards AD, Arthurs OJ. Paediatric MRI under sedation: is it necessary? What is the evidence for the alternatives? Pediatr Radiol 2011;41:1353–64.

2. Harned RK, Strain JD. MRI-compatible audio/visual system: impact on pediatric sedation. Pediatr Radiol 2001;31:247–50. 3. Carter AJ, Greer ML, Gray SE, et al. Reducing the need for anaesthesia in children. Pediatr Radiol 2010;40:1368–74. 4. Bartels M, Althoff RR, Boomsma DI. Anesthesia and cognitive performance in children: no evidence for a causal relationship. Twin Res Hum Genet 2009; 12(3):246–53. 5. Flick RP, Kutasic SK, Colligan RC, et al. Cognitive and behavioral outcomes after early exposure to anesthesia and surgery. Pediatrics 2011;128: e1053–61. 6. Hagmann P, Jonasson L, Madaer P, et al. Understanding diffusion MR imaging techniques: from scalar diffusion weighted imaging to diffusion tensor imaging and beyond. Radiographics 2006;26:S205–23. 7. Kocaoglu M, Bulakbasi N, Sanal HT, et al. Pediatric abdominal masses: diagnostic accuracy of diffusion weighted MRI. Magn Reson Imaging 2010; 28:629–36. 8. Humphries PD, Sebire NJ, Siegel MJ, et al. Tumors in pediatric patients at diffusion weighted MR imaging: apparent diffusion coefficient and tumor cellularity. Radiology 2007;245(3):848–54. 9. Gahr N, Darge K, Hahn G, et al. Diffusion-weighted MRI for differentiation of neuroblastoma and ganglioneuroblastoma/ganglioneuroma. Eur J Radiol 2011;79:443–6. 10. McDonald K, Sebire NJ, Anderson J, et al. Patterns of shift in ADC distributions in abdominal tumors during chemotherapy: a feasibility study. Pediatr Radiol 2011;41:99–106. 11. Padhani A, Koh DM, Collins DJ. Whole body diffusion-weighted MR imaging in cancer: current status and research directions. Radiology 2011; 261(3):700–18. 12. Kwee TC, Takahara T, Vermoolen MA, et al. Whole body diffusion-weighted imaging for staging malignant lymphoma in children. Pediatr Radiol 2010;40: 1592–602. 13. Kwee TC, Takahara T, Ochiai R, et al. Diffusionweighted whole-body imaging with background body signal suppression (DWIBS): features and potential applications in oncology. Eur J Radiol 2008; 18:1937–52. 14. Kwee TC, Quarles van Ufford HM, Beek FJ, et al. Whole-body MRI, including diffusion-weighted imaging, for the initial staging of malignant lymphoma: comparison to computed tomography. Invest Radiol 2009;44:683–90. 15. Daldrup-Link HE, Franzius C, Link TM, et al. Wholebody MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. Am J Roentgenol 2001;177:229–36.

Pediatric Abdominopelvic Oncologic MR Imaging 16. Mazumdar A, Siegel MJ, Narra V, et al. Whole-body fast inversion recovery MR imaging of small cell neoplasms in pediatric patients: a pilot study. Am J Roentgenol 2002;179:1261–6. 17. Laffan EE, O’Connor R, Ryan SP, et al. Whole-body magnetic resonance imaging: a useful additional sequence in paediatric imaging. Pediatr Radiol 2004;34:472–80. 18. Goo HW, Choi SH, Ghim T, et al. Whole body MRI of paediatric malignant tumours: comparison with conventional oncologic imaging methods. Pediatr Radiol 2005;35:766–73. 19. Goo HW. Whole body MRI of neuroblastoma. Eur J Radiol 2010;75:306–14. 20. Chavhan GB, Babyn PS. Whole-body MR imaging in children: principles, technique, current applications and future directions. Radiographics 2011;31: 1757–72. 21. Kellenberg CJ, Epelman M, Miller SF, et al. Fast STIR whole-body MR imaging in children. Radiographics 2004;24:1317–30. 22. Darge K, Jaramillo D, Siegel MJ. Whole-body MRI in children: current status and future applications. Eur J Radiol 2008;68:289–98. 23. Seigel MJ, Acharyya S, Hoffer FA, et al. Whole body MR imaging for staging of malignant tumors in pediatric patients: results of the American College of Radiology Imaging Network 6660 trial. Radiology 2013;266(2):599–609. 24. Antoch G, Voit FM, Freudenberg LS, et al. Wholebody dual modality PET/CT and whole-body MRI for tumor staging in oncology. JAMA 2003;290(24): 3199–206. 25. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases. Nature 2000;407:249–57. 26. van Dijke C, Brasch R, Roberts T, et al. Mammary carcinoma model: correlation of macromolecular contrast-enhanced MR imaging characterization of tumor microvasculature and histologic capillary density. Radiology 1996;198:813–8. 27. Daldrup H, Shames DN, Wedland M, et al. Correlation of dynamic contrast-enhanced magnetic resonance imaging with histologic tumor grade: comparison of macromolecular and small-

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

molecular contrast media. Pediatr Radiol 1998;28: 67–78. Padhani AR. Dynamic contrast-enhanced MRI in clinical oncology: current status and future directions. J Magn Reson Imaging 2002;16:407–22. Padhani AR, Husband JE. Dynamic contrastenhanced MRI studies in oncology with an emphasis on quantification, validation and human studies. Clin Radiol 2001;56:607–20. Tofts PS, Brix G, Buckley DL, et al. Estimating kinetic parameters from dynamic contrast-enhanced T1weighted MRI of a diffusible tracer: standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–32. Crone C. The permeability of capillaries in various organs as determined by the use of ‘indicator diffusion’ method. Acta Physiol Scand 1963;58: 292–305. Yuh WT. An exciting and challenging role for the advanced contrast MR imaging. J Magn Reson Imaging 1999;10:221–2. Taylor JS, Reddick WE. Evolution from empirical dynamic contrast-enhanced magnetic resonance imaging to pharmacokinetic MRI. Adv Drug Deliv Rev 2000;41:91–110. Kershaw L, Buckley DL. Precision in measurements of perfusion and microvascular permeability with T1weighted dynamic contrast enhanced MR. Magn Reson Med 2006;56:986–92. Reddick WE, Wang S, Xlong X, et al. Dynamic magnetic resonance imaging of regional contrast access as an additional prognostic factor in pediatric osteosarcoma. Cancer 2001;91:2230–7. Guo J, Reddick WE, Glass JO, et al. Dynamic contrast-enhanced magnetic resonance imaging as a prognostic factor in predicting event-free and overall survival in pediatric patients with osteosarcoma. Cancer 2012;118:3776–85. Kjaer A, Loft A, Law I, et al. PET MRI in cancer patients: first experiences and vision from Copenhagen. MAGMA 2013;26(1):37–47. Antoch G, Bockisch A. Combined PET/MRI: a new dimension in whole-body oncology imaging? Eur J Nucl Med Mol Imaging 2009;36:S113–20.

841